High-precision and remote-controlled methane and ethane detectors assist scientists in determining methane emission levels from natural gas development in north-central Texas’ Barnett Shale.

In October 2013, a pair of scientists took a technologically tripped-out van on a deliberately meandering methane detection mission across north-central Texas.

Tara Yacovitch and Scott Herndon of Massachusetts-based Aerodyne Research comprised one of 20 teams that came to Texas’ Barnett Shale to track leaking methane, a greenhouse gas that warms Earth‰’s atmosphere 84 to 87 times more than carbon dioxide within a 20-year time frame.

Previous to this extensive research expedition, data on methane leaks from natural gas development was often contradictory. The research teams of 2013 sought to gather a multitude of measurements that would finally nail down a consistent picture of the actual climate impacts of natural gas production.

Herndon drove as the Aerodyne Mobile Laboratory (AML) crisscrossed the Barnett Shale — one of the world’s most intensively developed natural gas fields and an early testbed for the hydraulic fracturing or “fracking‰” methods used to stimulate natural gas production. Yacovitch rode shotgun, monitoring as the AML’s instruments sniffed the air for leaking methane.

About every six or seven minutes, the readings climbed and fell as the AML moved through invisible plumes of methane. Meanwhile, Yacovitch checked wind direction and a map of local oil and gas facilities against what she saw out the window to pinpoint a given plume’s most likely source. Post-campaign analysis would integrate the AML’s stored data to estimate each source’s methane leak rates, which ranged widely from a low of 6 milligrams to a high of 571 kilograms per hour.

It was just one piece in a larger effort delivering a diversity of data. Aircraft flyovers generated top-down methane estimates for the region, for example, while methane detection via unmanned aerial vehicle (UAV) refined emissions estimates for a specific site. Other teams performed a detailed audit of leaks at a major gas compressor station.

The Barnett Shale campaign, in sum, constituted the most comprehensive regional examination of methane emissions ever undertaken. It also carried out science on an unprecedented scale for the organization coordinating it: the Environmental Defense Fund (EDF), a Washington, D.C.-based environmental advocacy group.

What inspired this unusual program was the need to quantify exactly how much methane oil and gas producers release. This stubborn question has complicated the work of environmental regulators and energy planners, obscuring the true environmental costs and benefits of burning natural gas and confusing efforts to gauge its proper role in future energy systems.

The Barnett Shale campaign — along with other recent research — appears to have closed the debate over whether methane escaping into the atmosphere exceeds official figures from the U.S. Environmental Protection Agency (EPA).

EPA acknowledged the gap in late February, releasing draft revisions to its Greenhouse Gas Inventory that put U.S. oil and gas industry emissions of methane at 9.3 million metric tons for 2013. That is 27 percent higher than EPA’s prior tally for the year.

Burning natural gas produces less carbon dioxide than burning coal or petroleum, making it a potentially climate-friendly alternative fuel for both power plants and vehicles. However, since methane is such a potent greenhouse gas, leaking just a little natural gas before it reaches a burner can offset the climate advantage promised by lower carbon dioxide (CO2) emissions. Indeed, the emissions measured from the Barnett natural gas supply chain represent a 50 percent increase in net radiative forcing.

For more than five years, Earth observation missions by airplanes and satellites — so-called “top-down” studies — have consistently detected more methane in the atmosphere than “bottom-up” estimates of known methane sources suggest should be there. The possibility of extra emissions is especially disturbing in light of the Paris Agreement reached in December, which commits the international community to holding total warming since the industrial revolution to 2 degrees C or, ideally, 1.5 C — targets that demand far greater emissions reductions than those presently contemplated.

Oil and gas wells are not the only potential cause — landfills and livestock operations also generate methane — but the top-down studies have increasingly pointed a finger at natural gas fields. A 2014 meta-analysis of top-down emissions measurements in the journal Science estimated that natural gas infrastructure was releasing 25 to 75 percent more methane than EPA registries recognized.

The Barnett Shale campaign sought to close this methane data gap by simultaneously studying one region with an array of top-down and bottom-up methods. The results, synthesized in a November 2015 report in the Proceedings of the National Academy of Sciences (PNAS), say that this scientific assault finally achieved closure for the study area: Top-down and bottom-up measures of methane leakage from natural gas systems were within 10 percent of each other, while the estimates of total methane emissions were within 0.1 percent.

“The top down and bottom up agreed remarkably well for the first time,” said Princeton University atmospheric chemist and engineering professor Mark Zondlo, one of the campaign’s principal investigators.

EDF’s results show that gas operations are releasing a lot more methane than EPA expected — about 90 percent more in the Barnett Shale. EPA’s Greenhouse Gas Inventory estimates oil and gas facilities in the 25-county region making up the Barnett Shale at 31 metric tons per hour, whereas the Barnett Shale estimated it to be 59 metric tons per hour.

They also showed how intermittent releases can be, and how methane pollution rates can be variable between sites that, on the surface, appear to be equivalent.

“We’ve gone beyond the point of saying, ‘Well this method said this and that method said that.’ The methods are now solid enough to show that there’s a problem here that needs to be addressed,” said Zondlo.

Mark Boling, senior vice president for Houston-based gas producer Southwestern Energy, agrees that the Barnett campaign makes it tough for the gas industry to dismiss methane leakage. Boling said he sees “much fewer” people in his industry denying the problem’s existence than just six months ago.

“People are starting to recognize that methane emissions are a legitimate issue that needs to be dealt with,” said Boling.

EDF developed the Barnett campaign based on input from academic and government scientists such as Zondlo, then secured a $1 million grant from the Alfred P. Sloan Foundation to finance it, and launched the campaign 12 months after it was conceived.

Much of the equipment and methodology deployed was state-of-the-art. Aerodyne Research brought its first high-precision mobile ethane detector to the campaign. The detector provided ethane measurements from the Aerodyne Mobile Laboratory van and Barnett campaign aircraft flyovers.

Measuring methane and ethane simultaneously enabled researchers to distinguish methane plumes associated with fossil fuel production from methane belched by livestock and seeping from landfills, known as biogenic methane. While natural gas contains ethane along with its methane, biogenic methane is ethane-free.

Aerodyne’s Yacovitch and Herndon used this new capability during their expedition crisscrossing the Barnett Shale. Yacovitch said the telltale signal of fossil methane was plain as day.

“We were moving and seeing methane levels rise and fall and seeing ethane rise and fall with it,” she recalled.

The UAV-based methane measurement used by another team, for aerial measurements, also was a first-of-a-kind, according to Zondlo, who collaborated with David Lary of the University of Texas to build a high-precision methane detector just one-tenth the size and weight of commercial detectors — small enough to fit on a model airplane with a 3-meter wingspan.

‰”We used some laser spectroscopic techniques to do this on a package that’s a couple of kilograms and a couple of liters,” said Zondlo.

They used their remote-controlled methane detector to circumnavigate a natural gas installation several times within 7-8 minutes, measuring plumes that were lofted high beyond the view of the mobile laboratory vans. These repeated measurements also revealed the degree to which emissions rates could rapidly shift.

The methods, which delivered high quality data from outside oil and gas facilities, enabled the campaign teams to collect data on hundreds of sites during the two-week Barnett-Shale campaign. The work also was less reliant on industry cooperation, avoiding a potential selection bias against low-performing sites and making the campaign less susceptible to accusations of industry influence that anti-fracking activists have leveled at some other EDF-led methane studies.

Most of the campaign teams from the Barnett Shale study reported on their pieces for solving the emissions puzzle in a special July 2015 issue of the journal Environmental Science & Technology. But definitive answers to the methane mystery were delivered in the November 2015 PNAS report, coauthored by several of the campaign’s principal investigators and EDF scientists, which integrated the array of findings to turn the highly variable data into accurate estimates of total methane leakage.

“We were able to bring multiple datasets together to build the first really robust statistical models,” said Steven Hamburg, EDF’s chief scientist.

The synthesis calculates that the Barnett’s natural gas supply chain — from the wellhead to the gas pipelines — fumbles 1.5 percent of the fossil methane that it liberates. That takes quite a bite out of the climate benefit of natural gas, according to Hamburg.

“The emissions that were measured from the Barnett natural gas supply chain represent a 50 percent increase in net radiative forcing — the amount of impact that burning natural gas has on the climate — over the first 20 years post combustion,” said Hamburg.

What does that mean for energy policy? According to the PNAS paper, it makes natural gas from the Barnett Shale a losing option for fueling heavy trucks, from a greenhouse gas perspective. Using compressed gas to wean heavy trucks off of diesel fuel would increase their impact on climate change for at least several decades.

The policy message is more nuanced regarding the role of natural gas on the power grid. Burning Barnett gas to generate electricity is preferable to burning coal from a climate perspective, according to the synthesis paper. But that might not hold true for gas from other regions, where top-down studies suggest methane leakage rates are higher than the Barnett’s. A 2014 report on Colorado’s Denver-Julesburg Basin, for example, documented methane emissions from oil and gas production nearly three times higher than those recognized by EPA.

In early 2015, the Obama administration set a goal of cutting methane emissions from the oil and gas sector by 40 to 45 percent from 2012 levels by 2025. To that end, the EPA will complete regulations in 2016 that mandate the use of better equipment at new gas production sites. For the more than 500,000 wells already producing natural gas in the U.S., the administration is emphasizing voluntary industry action.

The debate now turns to how to reduce those emissions. There is a lot the industry can do, according to Boling at Southwestern Energy, since the Barnett campaign shows that a minority of sites — so-called “super-emitters” — cause most of the problem.

“Something like 2 percent of the emissions sources were responsible for 50 percent of the emissions and just 10 percent of the facilities were responsible for 90 percent of the emissions,” noted Boling, “If you want to get the highest reduction in methane emissions with the least amount of investment, you really have to find where these super-emitters are and deal with them.”

Boling says his company is ramping up its efforts to find its own super-emitters and fix them. The firm has acquired infrared methane detection cameras to spot leaks, as well as the methane measurement devices to measure the flows. And they are leading players in EPA’s Methane Challenge Program, a voluntary partnership with industry.

EPA also is working to clean up its numbers. In a statement prior to its recent data revision, an agency spokesperson confirmed that it was “reviewing the EDF study on methane emissions from natural gas systems in the Barnett Shale,” and that it was one of the inputs to an ongoing effort to “refine its emission estimates.”

The agency has hired one of the Barnett campaign’s scientific advisers — Harvard University environmental engineer Daniel Jacob — to help find and fix the holes in its official registry of methane emissions. Jacob says his project is using the Barnett Shale data as a benchmark for improving, for example, how EPA’s inventory accounts for super-emitters.

“We’re using the Barnett Shale experiment to actually define and quantify the errors in EPA’s National Inventory,” said Jacob.

Zondlo says transforming the Barnett Shale’s results to impact government policy and industry practice was one of the reasons he appreciated being part of EDF’s project. He attributes the speed to the fact that EDF was plugged into the policy arena at the state and federal levels and, as an advocacy organization, it was explicitly oriented toward agitating for action.

“That was the fun of the campaign for me,” said Zondlo. “I knew it was going to become relevant immediately.”

Peter Fairley is an environment and energy journalist who has contributed to Earthzine since its launch in 2007. Fairley profiled the growing role of top-down measurements of greenhouse gas emissions as a check on bottom-up inventories in a 2012 Earthzine feature.